This documentation explains how to set up the Teams Transcripts Connector with the Abacus.AI enterprise application. If you have already set up the connector, skip to How to Use the Teams Transcripts Connector.
To integrate Microsoft Teams with Abacus.AI and enable the Teams Transcripts Connector, follow these steps:
Accept
to grant permission to access Teams data.
Once the Teams Transcripts Connector is set up, you can use it to create datasets for training models in Abacus.AI.
Create a new project and select the use case, then go to the "Datasets" tab and click "Create Dataset".
Name the dataset, choose "Type of Data: Tabular data", then click "Continue".
Choose "Read from External Service" and select your Microsoft Teams connector under "Application Connectors".
With your connector selected, toggle the options for the data you want to pull: - Meeting Transcripts: Pull transcripts from meetings with transcripts enabled. Transcripts are accessible only for meetings not associated with channels. - Posts in Channels: Pull all posts across all channels that the user has access to. - Chat Messages: Pull all messages from one-to-one and group chats that the user has access to.
After selecting the data, click on "Select" and then "Add Dataset" (unless you want to perform advanced options like filtering or preprocessing).
Once the dataset is uploaded, configure the schema mapping and train models with the data.
When setting up the connector, the following permissions are requested:
- User.ReadBasic.All: Access to basic user information such as name and email.
- Team.ReadBasic.All: Access to basic team information.
- ChannelMember.Read.All: Access to channel member lists.
- Group.Read.All: Access to group information.
- ChannelMessage.Read.All: Access to channel messages.
- Calendars.Read: Access to read calendar schedules.
- OnlineMeetingArtifact.Read.All: Access to meeting artifacts like attendance reports.
- OnlineMeetings.Read: Access to meeting information.
- OnlineMeetingTranscript.Read.All: Access to meeting transcripts.
These permissions allow the connector to pull necessary data for analysis and model training.
The Teams Transcripts Connector enables you to:
What data is pulled by the connector?
The connector pulls meeting transcripts, chat messages, and posts in channels. It does not pull video or audio
recordings.
How are transcripts pulled?
Transcripts are pulled using the Microsoft Graph API.
Is transcription automatic for all meetings?
Transcription must be enabled by someone present at the meeting. It is not automatic unless configured by the user. If a service account is added to a meeting, the transcript would only be stored for the meeting after the point the service account has joined and transcription has started.
What is the retention policy for transcripts?
Transcripts are retained according to Microsoft's policy, typically for 60 days after the meeting ends.
Can we use our own transcription service?
Currently, the connector does not support using an external transcription service.
How long does synchronization take?
Synchronization occurs every 30 minutes, with data retained for up to 12 hours after deletion in Teams.
If you encounter any issues while setting up or using the Teams Transcripts Connector, refer to the following tips:
- Data Not Syncing: Verify that the scope is correctly configured and that the necessary permissions have been granted.
- Meeting Transcripts Missing: Ensure that transcription is enabled in your Teams meetings.
For further assistance, contact Abacus.AI support.